SVD for Very Large Matrices: An Approach with Polar Decomposition and Polynomial Approximation.
Masaki OnukiYuichi TanakaPublished in: ICDM Workshops (2018)
Keyphrases
- polynomial approximation
- singular value decomposition
- singular values
- pointwise
- least squares
- reconstruction error
- decomposition method
- principal component analysis
- fourier transform
- decomposition methods
- low rank
- low rank approximation
- lower bound
- data matrix
- rotation invariant
- frequency domain
- facet model
- dimensionality reduction
- positive definite